Workplace AI and Human Flourishing

This Article explores the important but largely unexplored relationship between workplace artificial intelligence (AI) and human flourishing. More specifically, the Article examines the potential impact of workplace AI decision tools on such critical matters as workers’ human dignity, workplace and personal autonomy, and the opportunity for upward mobility. AI can analyze data far more quickly and efficiently than humans. Moreover, AI computer models are far superior to people in uncovering subtle correlations in large amounts of data and learning from those correlations. Thus, workplace AI decision tools teach themselves to choose the criteria for recruitment, hiring, compensation, promotion, and termination of workers and ground their employment decisions firmly in empirical data that humans fail even to perceive. At the same time, these AI tools may insult human dignity, threaten worker autonomy, and serve as an agent of social ossification. This Article demonstrates that both efficiency and elements of human flourishing are important values in the law governing American workplaces. Thus, both should be given significant weight in formulating the regulation of workplace AI decision tools. The Article then proposes a means of balancing the interests of employers (and society more generally) in workplace efficiency against the need to protect and promote workers’ interests in human dignity, autonomy, and the opportunity for upward mobility. In particular, the Article argues that when an employer operates a factor-selecting AI decision tool on a worker that results in an adverse employment action with respect to that worker, the employer should bear the burden of demonstrating that the AI tool has only used factors that appear, on their face, relevant to the job at issue